regions and may indeed have gone much further than can be measured by reliable data. Second, in some cases the reduction of the birth rate in the near future may be slowed or even temporarily reversed by prospective changes in the age and sex structure of the countries concerned.* In a number of less developed countries there was a relatively low birth rate during the Depression and during the Second World War, followed by some recovery and a higher birth rate after the War. As a result, larger cohorts of young people are now moving into the reproductive ages than was the case during the mid-1960's. In Taiwan, for example, it will require a considerable acceleration in the reduction of age-specific fertilityt to continue the present trend toward a lower crude birth rate (10). Obviously one would like to have better measures of natality than the crude birth. Reproduction rates, which eliminate most of the variance caused by changing proportions of women in the marriageable and reproductive ages, would provide a better measure of trends. Unfortunately these statistics require accurate information on births by age of mother and on age distribution of women in the reproductive ages—information that is often unavailable for recent years. Where it is available, it shows downward trends paralleling those for crude birth rates. $ The available data for the 1960's support the reasonable conclusion that, given time and socioeconomic advance, one may expect a general diffusion of lower birth rates in the less developed countries. Clearly the process has begun. But much depends on how fast these reductions will occur and at what level of socioeconomic development they may be expected to begin. On the first point, do declines in natality, like declines in mortality, accelerate over time? THE ACCELERATING REDUCTION OF NATALITY IN THE DEMOGRAPHIC TRANSITION The death rate in the less developed areas is dropping very rapidly—"a decline that looks almost vertical compared to the gradual decline in Western *Sec Nathan Key fit?,, "Changes in Birth and Death Rates and Their Demographic Effects," in this volume. t Births per 1,000 women at each age, i.e., 15 to 19, 20 to 24, etc. $E.g., for Taiwan, Singapore, Chile, Costa Rica, Puerto Rico, Trinidad and Tobago, and Mauritius. Data on completed fertility, i.e., the total number of children born to women passing through their reproductive years, is the final and most accurate measure of fertility, but it docs not measure recent change, which is precisely what we are after here. One cannot get around the basic fact that women have births at changing intervals and over a long period of time. A woman at age 45, who has completed her childbearing, usually had her first child 20 to 30 years previously and her number of children reflectstions.